g protein-coupled estrogen receptor-selective ligands modulate

Hindawi Publishing CorporationObstetrics and Gynecology InternationalVolume 2013 Article ID 472720 17 pageshttpdxdoiorg1011552013472720

Research ArticleG Protein-Coupled Estrogen Receptor-SelectiveLigands Modulate Endometrial Tumor Growth

Whitney K Petrie1 Megan K Dennis1 Chelin Hu1 Donghai Dai23 Jeffrey B Arterburn4

Harriet O Smith125 Helen J Hathaway1 and Eric R Prossnitz1

1 Department of Cell Biology and Physiology and UNM Cancer Center The University of New MexicoHealth Sciences Center Albuquerque NM 87131 USA

2Department of Obstetrics and Gynecology UNM Cancer Center The University of New MexicoHealth Sciences Center Albuquerque NM 87131 USA

3Department of Obstetrics and Gynecology University of Iowa Carver College of Medicine Iowa City IA 52242 USA4Department of Chemistry and Biochemistry New Mexico State University Las Cruces NM 88003 USA5Department of Obstetrics and Gynecology and Womenrsquos Health Albert Einstein College of Medicine and MontefioreMedical Center Bronx NY 10461 USA

Correspondence should be addressed to Eric R Prossnitz eprossnitzsaludunmedu

Received 6 August 2013 Accepted 17 September 2013

Academic Editor Andrew P Bradford

Copyright copy 2013 Whitney K Petrie et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Endometrial carcinoma is the most common cancer of the female reproductive tract GPERGPR30 is a 7-transmembranespanning G protein-coupled receptor that has been identified as the third estrogen receptor in addition to ER120572 and ER120573High GPER expression is predictive of poor survival in endometrial and ovarian cancer but despite this the estrogen-mediatedsignaling pathways and specific estrogen receptors involved in endometrial cancer remain unclear Here employing ER120572-negative Hec50 endometrial cancer cells we demonstrate that GPER mediates estrogen-stimulated activation of ERK andPI3K via matrix metalloproteinase activation and subsequent transactivation of the EGFR and that ER-targeted therapeuticagents (4-hydroxytamoxifen ICI182780fulvestrant and Raloxifene) the phytoestrogen genistein and the ldquoER120572-selectiverdquo agonistpropylpyrazole triol also function as GPER agonists Furthermore xenograft tumors of Hec50 cells yield enhanced growth withG-1 and estrogen the latter being inhibited by GPER-selective pharmacologic antagonism with G36 These results have importantimplications with respect to the use of putatively ER-selective ligands and particularly for the widespread long-term use of ldquoER-targetedrdquo therapeutics Moreover our findings shed light on the potential mechanisms of SERMSERD side effects reported inmany clinical studies Finally our results provide the first demonstration that pharmacological inhibition of GPER activity in vivoprevents estrogen-mediated tumor growth

1 Introduction

Carcinoma of the endometrium is the most common cancerof the female reproductive tract with over 40000 newdiagnoses and over 7000 deaths per year in the UnitedStates Although the majority (sim75) of endometrial tumorsare of endometrioid histology (designated type I tumors)expressing high levels of estrogen receptor (ER) progesteronereceptor (PR) and epidermal growth factor receptor (EGFR)

about 25 of tumors are of advanced stage (designatedtype II tumors) are unlikely to be ER+PR+ and have apoorer prognosis [1] Although overlap exists with respectto histology genetic aberations and epidemiological profilesthe two tumor types appear to represent discrete carcinogenicprocesses with distinct molecular characteristics Type Itumors consist of well-differentiated tumors preceded byendometrial hyperplasia and are associated with a loss ofPTEN expression as well as abnormalities in 120573-catenin Kras

2 Obstetrics and Gynecology International

and DNA mismatch repair genes Type 2 tumors are a het-erogeneous group of tumors including high-grade (undiffer-entiated) endometrioid carcinomas uterine papillary serouscarcinomas clear cell carcinomas and carcinosarcomas withdifferent mutational profiles Over 90 of uterine papillaryserous carcinomas are associated with p53 mutations 45ndash60 have Her-2neu mutations and PTEN mutations arerare [2 3] Carcinosarcomas which are characterized bymalignant epithelial andmesenchymal components are asso-ciated with many of the epidemiological risk factors linkedto endometrioid carcinomas including obesity and exposureto tamoxifen therapy suggesting that dysregulated estrogensignaling has a role in their pathogenesis and may representa therapeutic target Moreover recent mutational profilingstudies indicate that whereas some carcinosarcomas sharemutations with type 1 tumors (PTEN and ARID1A) othersshare mutations with uterine papillary serous carcinomas(notably p53 and PPP2R1A) [4]

The lack of estrogen receptor 120572 (ER120572) expression inmost type II tumors has led to the assumption that thesetumors must be ldquoestrogen-independentrdquo and that treatmentwith antiestrogens (selective estrogen receptor modulators(SERMs) such as tamoxifen and Raloxifene or pure antag-onistsselective estrogen receptor modulators (SERDs) suchas ICI182780fulvestrant) commonly used in breast can-cer treatment would be ineffectual a conclusion largelysubstantiated in a number of clinical trials [5 6] In factprolonged treatment of breast cancer with SERMs such astamoxifen leads to an increased incidence of endometrialcancer [7] particularly those of high-risk histologic types[8] resulting in significantly poorer overall survival [9] Theeffects of tamoxifen in the uterus have been ascribed toaltered expression of nuclear coregulatory proteins in theendometrium compared to the breast resulting in moderateagonist activity of tamoxifen in the uterus compared to itsantagonistic effects in the breast [10ndash13] However recentresults have suggested that a heretofore-underappreciatedestrogen receptor the G protein-coupled estrogen recep-tor (GPER formerly GPR30) may play an important rolein both the increased incidence of endometrial cancer inwomen treated with tamoxifen [14] as well as representingan alternate mechanism through which endometrial cancersparticularly type II tumors can maintain responsiveness toestrogen [15]

GPER is a member of the 7-transmembrane spanningG protein-coupled receptor (GPCR) superfamily struc-turally unrelated to the nuclear receptor family membersER120572 and ER120573 [16 17] Activation of GPER by estrogenhas been demonstrated in many cancer cell lines [18 19]including endometrial cancer cells [15 20ndash27] GPER isalso activated by antiestrogens including tamoxifen (ie 4-hydroxytamoxifen) [28] and ICI182780 (fulvestrant) [29]leading to the suggestion that GPER plays a role in hormone-resistance in breast cancer [30 31] as well as in theincreased incidence of endometrial cancer in women takingtamoxifen for breast cancer [14 32] Furthermore GPER(over)expression has been associated with many cancersand in particular poor prognosis in a number of cancersincluding breast [33] ovarian [34] lung [35] pancreatic [36]

and endometrial [37] although observations to the contraryhave also been reported [38 39]

Because of the lack of specificity of estrogen and anti-estrogens for the three known estrogen receptors (ER120572 ER120573andGPER)we have developed both aGPER-selective agonist(G-1 [40]) and antagonists (G15 [41] and G36 [42]) thatdisplay virtually no activity towards the classical estrogenreceptors In the current study we examine the expressionand function of GPER in the ER120572minusER120573minus endometrial cancercell line Hec50 [43 44] which is representative of type IIendometrial cancers We demonstrate that in Hec50 cellsGPER is localized predominantly in intracellular membranesand mediates PI3K and ERK activation in response to bothestrogen and G-1 as well as the antiestrogens tamoxifenRaloxifene and ICI182780 and the ldquoER120572-selectiverdquo agonistpropylpyrazole triol (PPT) We also demonstrate that Hec50cells and primary patient endometrial adenocarcinomasmaintain expression of GPER when grown as xenografttumors Finally we demonstrate both estrogen and G-1stimulate Hec50 xenograft tumor growth in vivo and that theGPER antagonist G36 greatly reduces growth of estrogen-stimulated Hec50 tumors Overall these results suggest thatGPER may play a critical role in endometrial carcinogenesisproviding a novel target for prognosis and treatment

2 Materials and Methods

Reagents 17120573-estradiol 17120572-estradiol 4-hydroxytamoxifenRaloxifene genistein LY294002 bovine serum albumin(BSA) normal goat serum insulin transferrin hydrocor-tisone fetuin pancreatin and trypsin were from Sigma(St Louis MO USA) AG1478 and GM6001 were fromCalbiochem (BillericaMA USA) DPN PPT and ICI182780were from Tocris Chemicals (Ellisville MO USA) G-1G15 and G36 were synthesized as previously described [40ndash42] Goat anti-rabbit Alexa-488 goat anti-rabbit Alexa-568and donkey anti-mouse Alexa-568-conjugated secondaryantibodieswere from Invitrogen (Carlsbad CAUSA) Rabbitanti-GPER C-terminal antiserum (cross-reactive to bothhuman and murine sequences) was produced as previouslydescribed and used at a dilution of 1 10000 [28] Rab-bit anti-GPER human N-terminal antiserum was producedagainst the peptide sequence MDVTSQARGVGLEMYPG-TAQPAAC (with an added carboxy-terminal cysteine forconjugation to KLH) by New England Peptide Inc (GardnerMA USA) and used at a dilution of 1 5000 Polyclonalantibodies against total ERK and pERK were from Cell Sig-naling (Danvers MA USA) monoclonal anti-actin and anti-120573-catenin antibodies were from Millipore (Burlington MAUSA) Goat anti-rabbit HRP and donkey anti-mouse HRPwere from GE-Amersham (Piscataway NJ USA) DulbeccorsquosMEM RPMI 1640 and phenol red-free DMEMF12 mediapenicillinstreptomycinglutamine and fetal bovine serumwere obtained from Fisher (Pittsburgh PA USA)

Cell Culture andTransfection Human endometrial carcinomaIshikawa H cells and Hec50 (specifically Hec50co [44]) cells(kindly provided by K K Leslie) were cultured in DMEMmedium with FBS (10) and 100UmL penicillin 100 120583gmL

Obstetrics and Gynecology International 3

streptomycin and 2mm L-glutamine Cells were grown asa monolayer at 37∘C in a humidified atmosphere of 5CO2and 95 air PH-RFP plasmid DNA was transfected

with Lipofectamine2000 according to manufacturerrsquos direc-tions but using 14 the recommended amount of DNA Forldquoco-transfectionrdquo of PH-RFP plasmid DNA and siRNA oncoverslips Lipofectamine2000 was used to transfect siRNAon day 1 according to manufacturerrsquos directions on day2 PH-RFP plasmid with additional siRNA as on day 1was retransfected For GPER knockdown siRNA targetingGPER (ON-TARGET plus SMARTpool L-005563-00) wasobtained from DharmaconThermo-Fisher (Lafayette COUSA) The nontargeting siRNA ON-TARGETplus siControlNon-Targeting siRNA (D-001810-02) was used as a controlCells transfected with siRNA were used in PH assays andstained for GPER expression 48 hours following the secondsiRNA transfection For microscopy experiments cells wereseeded onto 12mm glass coverslips and allowed to adherefor at least 24 h prior to antibody staining or 12 h prior totransfection

Primary mouse uterine epithelial cells were harvestedand cultured as described [45] Briefly uteri were removedfrom C57Bl6 mice between 21 and 35 days of age and slitlongitudinally followed by incubation in 25 pancreatin and05 trypsin for 1 h at 4∘C followed by 1 h at RT Digesteduteri were then briefly vortexed releasing epithelial sheetsand fragments which were transferred to a fresh tube con-taining 2 trypsin inhibitor (Invitrogen) in Hanks BalancedSalt Solution (HBSS) After two washes with HBSS cells wereseeded directly onto acid-washed poly-L-lysine-coated glasscoverslips and cultured in serum-free DMEMF12 mediumsupplemented with 5 120583gmL insulin 10 120583gmL transferrin10minus7M hydrocortisone 2mgmL BSA 1mgmL fetuin and

antibiotics

Immunofluorescence Staining Cells were seeded on 12mmglass coverslips and fixed with 4 PFA (ElectronMicroscopySciences Hatfield PA USA) in PBS for 15min at 37∘C Cov-erslips were washed three times with PBS and blocked for 1 hwith 3 BSA in PBS Where indicated permeabilization wasaccomplished with 005 Triton X-100 in the blocking bufferPrimary antibody was diluted in 3 normal goat serumand coverslips were incubated for 4 h at room temperatureCoverslips were washed three times with PBS and incubatedwith secondary antibody diluted in 3 normal goat serumCoverslips were washed three times with PBS and mountedwith Vectashield containing DAPI (Vector Labs BurlingameCA USA) Confocal fluorescence images were collected ona Zeiss LSM 510 confocal microscope Typical cell lengthsranged from 20 to 35 120583m

Western Blotting Cells were harvested directly for receptorexpression or starved in phenol red-free RPMI 1640 for 24 hprior to treatment Cells were washed once with ice-coldPBS and lysed using NP-40 buffer Twenty 120583g protein wasloaded per lane and electrophoresed on 4ndash20 SDS-PAGEgels (Thermo Scientific Waltham MA USA) transferred toPVDF membrane (Millipore) and blocked with 3 BSA inTBST (50mmTris 150mmNaCl and 01 Tween-20) before

overnight incubation with primary antibodies at 4∘C Blotswere incubated with HRP-conjugated secondary antibodiesdeveloped using SuperSignal West Pico ChemiluminescentSubstrate (Thermo Scientific) exposed to film scanned andquantified using Image J software (NIH)

PI3K Activation The pleckstrin homology (PH) domain(responsible for PIP3 binding) of Akt [46] fused tomonomeric red fluorescent protein 1 (mRFP1) [47] generatedthe PH-RFP construct which was used to localize sites ofincreased cellular PIP3 accumulation [28] Hec50 cells wereplated on coverslips transfected with PH-RFP followed bya 24 h recovery in complete DMEM and serum starvedin phenol red-free DMEMF12 for 24 h before stimulationwith ligands as indicated The cells were fixed with 2 PFAin PBS washed mounted in Vectashield containing DAPI(Vector Labs) and imaged by confocal microscopy using aZeiss LSM510 confocal fluorescence microscope Images arerepresentative of 75ndash85 of the transfected cells observed

Xenograft Tumors Ishikawa H cell and Hec50 cell xenografttumors were produced by injecting sim 3 times 106 cells (in 100120583LDMEM) subcutaneously into 6ndash8-week-old female athymicCrlNuNu-nuBR ldquoathymic nuderdquo mice [48] Subcutaneoustumors were recovered for histology and immunohistochem-istry typically sim6 weeks after injection when the tumorsreached sim10mm in diameter For treatment models Hec50cells were injected into ovariectomized athymic nude mice10 days after ovariectomy Individual 60-day-release shamestrogen (15mg) G-1 (225mg (equimolar with estrogen))and G36 (11mg pellets (5-fold molar excess versus estrogenand G-1)) custom made by Innovative Research of America(Sarasota FL) were introduced subcutaneously near thescapula with a trochar on the same day as Hec50 cellinoculation When tumors became palpable (3ndash5mm indiameter) tumor size was measured sim3 times per week bydigital caliper and upon sacrifice tumors were dissected andweighed All protocols were approved by the InstitutionalAnimal Care and Use Committee of the University of NewMexico Health Sciences Center

For xenografts of fresh human tumor samples tumorswere obtained from the Department of Pathology at theUniversity of New Mexico immediately after their arrival atthe Surgical Pathology Gross Room Fat and necrotic tissueswere trimmed and remaining tumor tissues were rinsed withcell culture medium (DMEM) Tumors were minced into afine homogenate and mixed with medium Typically 10mgof tumor tissue was mixed with 100 uL medium for subcu-taneous injection into a 6ndash8-week old athymic CrlNuNu-nuBR female mouse to create the first-generation xenograftswhich were used for analysis as reported previously [49]Thecollection of human endometrial tumors was approved byHuman Research Protections Office at the University of NewMexico

Histological Staining of Tumors Five-micron sections fromparaffin-embedded tumor tissues were prepared for im-munohistochemistry (IHC) as previously described usingthe carboxy-terminus-targeted antibody against GPER [37]


Hec50

H cell

Control siRNA GPER siRNA

(a)

GPER

Actin

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(b)

lowast

Relat

ive e

xpre

ssio

n (

)

0

20

40

60

80

100

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(c)

Figure 1 GPER expression in type I and type II endometrial cancer cells (a) Representative immunofluorescence images of Ishikawa H cells(type I) and Hec50 cells (type II) transfected with either control siRNA or siRNA targeting GPER GPER is shown in green nuclei are stainedblue with DAPI (b) Western blot of GPER and actin in untransfected control siRNA- and GPER-targeted siRNA-transfected H cells andHec50 cells (c) Western blot quantitation of GPER expression relative to actin and normalized to untransfected Hec50 cells Data representmean plusmn sem from three experiments lowast119875 lt 005 compared to control siRNA

In brief sections were deparaffinized in CitriSolv clearingagent (Fisher Pittsburgh PA USA) followed by rehydrationin increasing H

2O ethanol solutions Antigen retrieval was

accomplished bymicrowaving slides in 001M sodium citratebuffer (pH 60) for 25min followed by incubation of cooledslides in fresh 2 H

2O2for 10min Permeabilization and

blocking were performed by incubating the slides for 30minin 200120583L of 01 Triton X-100 in PBS with 3 bovine serumalbumin in a humid chamber Slides were incubated with theaffinity-purified GPER carboxy-terminal antibody diluted toa final protein concentration of 2120583gmL in 3 normal goatserum for 1 h Following multiple washes bound antibodywas detected using the immunoperoxidase system by incu-bating with goat anti-rabbit IgG conjugated to horseradishperoxidase (diluted 1 250 in 3 normal goat serum) for45min Peroxidase was detected with the enzyme substrate

310158403-diaminobenzidine tetrahydrochloride (DAB Sigma StLouis MO USA)

3 Results

31 GPER Is Expressed Intracellularly in Hec50 Type IIEndometrial Cancer Cells Although the physiological andbiological effects of estrogen have traditionally been de-scribed as being mediated by the nuclear estrogen receptorsER120572 and ER120573 recent evidence suggests an increasing rolefor the 7-transmembrane estrogen receptor GPER [16] Wehave previously observed that in many cell types stainingfor GPER reveals a predominantly intracellular pattern asso-ciated with the endoplasmic reticulum and Golgi apparatusIntracellular localization including the nucleus [50 51] has


Permeabilized Unpermeabilized

N-Term

C-Term

GPER 120573-catenin

Figure 2 Subcellular localization of GPER expression in pri-mary mouse uterine epithelial cells Cells were stained with eitheran antibody targeted against the carboxy terminus (C-Term) oramino terminus (N-Term) of GPER under either permeabilizing ornonpermeabilizing conditions As the amino terminus of GPCRswould be extracellular any plasma membrane-localized receptorshould be stained by the amino terminus-targeted antibody undernonpermeabilizing conditions GPER is stained green 120573-catenin isstained red as a plasma membrane marker Arrow indicates nuclearmembrane

been reported by many [52ndash55] but not in other studies [56ndash58] Recent results suggest that GPER undergoes constitutiveinternalization which would suggest that at steady state apreponderance of GPER would be detected as intracellular[59 60] To address this further in endometrial epithelialcancer cells we examined expression in type II Hec50 cellsand compared it to type I Ishikawa H cells Immunofluores-cence staining andWestern blotting (both with an anti-GPERcarboxy terminus-targeted antibody) revealed that IshikawaH cells express very low levels of GPER (sim10) compared toHec 50 cells (Figures 1(a) and 1(b)) Treatment of both celltypes with siRNA resulted in a significant reduction of GPERexpression in Hec50 cells with no significant reduction in Hcells (Figure 1) Immunofluorescence staining also revealed apattern of GPER staining throughout the cytoplasm consis-tent with localization to intracellular membranes

To assess the localization of GPER in greater detail inprimary cells we cultured freshly isolated mouse uterineepithelial cells on glass coverslips and stained for both thecarboxy-terminus as well as the amino-terminus of GPERin combination with 120573-catenin as a cell surface markerunder both permeabilizing and nonpermeabilizing condi-tions (Figure 2) Staining for GPER with both antibodiesunder permeabilizing conditions revealed an intracellularlocalization with no significant overlap with 120573-catenin Inter-estingly in many cells (eg upper left panel) there was alsostaining of the nuclear membrane which is continuous withthe endoplasmic reticulum Since GPCRs are oriented inthe plasma membrane with their amino terminus to the cellexterior and their carboxy terminus to the cell cytoplasmwe expected that if GPERs were expressed in the plasmamembrane staining with an antibody targeting the aminoterminus should be able to detect any receptor in the plasma

membrane in the absence of permeabilization Howeverstaining with the amino terminus-targeted antibody in theabsence of permeabilization revealed no significant stainingin contrast to the staining observed following permeabi-lization indicating that little GPER is expressed on the cellsurface compared to the intracellular pool

32 PI3K Activation by Estrogen in Hec50 Endometrial Can-cer Cells Is Mediated by GPER As Hec50 cells lack ER120572expression [44] we next asked whether in the absence ofER120572 estrogen could still mediate rapid signaling To addressthis we utilized a method we have previously employed[28 41 42] monitoring the activation of PI3K through thetranslocation of a fluorescent reporter of PIP3 localizationnamely a red fluorescent reporter (mRFP1 [47]) proteinfused to the PH (PIP3-binding) domain of Akt [46] Inprevious studies we have observed that in serum-starvedunstimulated cells (eg COS7 cells) the PH-RFP reporteris fairly uniformly distributed throughout the cytoplasm andnucleus [28] However when expressed in certain cancer celllines (eg SKBr3 breast cancer cells) the reporter exhibitsan enhanced plasmamembrane localization [28] even underserum-starved conditions that is likely due to constitutiveactivation of signaling pathways (eg EGFR activation orHer2 overexpression) that lead to activation of PI3K in theabsence of exogenous stimuli

To determine whether estrogenmediates rapid activationof PI3K in Hec50 cells we transfected cells with PH-RFPand subsequently treated serum-starved cells with estrogen(17120573-estradiol) The unstimulated cells yielded a plasmamembrane localization similar to that previously observed inSKBr3 breast cancer cells suggesting constitutive activationof PI3K at the plasma membrane (Figure 3(a)) Howeverupon estrogen stimulation for 15min the reporter translo-cated to the nucleus suggesting activation of PI3K in thenucleus as has been suggested by studies characterizinga nuclear pool of PI3K [61 62] Importantly the inactivestereoisomer of estrogen (17120572-estradiol) did not demonstratePI3K activation even at 1000x the concentration of 17120573-estradiol demonstrating the stereoselectivity of the receptorinvolved for the physiologically active isomer of estrogen Totest whether the activity of PI3K was required for the translo-cation of the PH-RFP reporter we pretreated cells with thePI3K inhibitor LY294002 followed by estrogen stimulationwhich yielded a uniform distribution of the PH-RFP reporterthroughout the cell This indicated not only that the nuclearlocalization of the PH-RFP reporter required PI3K activitybut that the membrane localization in unstimulated cellswas also due to PI3K activity (as LY294002 treatment in theabsence of estrogen yielded an identical distribution data notshown) As rapid estrogen signaling has been demonstratedto involverequire EGFR activation through the generationof HB-EGF [29] we also pretreated cells with the EGFRkinase inhibitor AG1478 and the metalloproteinase inhibitorGM6001 (to block HB-EGF production) Both inhibitorsblocked estrogen-mediated nuclear accumulation of the PH-RFP reporter indicating a requirement for both HB-EGF


Unstimulated 10nM 17120573E2 10120583M 17120572E2

17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)

Figure 3 Activation of PI3K by estrogen in Hec50 cells is mediated by GPER Hec50 cells were transfected with amarker of PIP3 productionthe PH domain of Akt fused to monomeric red fluorescent protein (mRFP) yielding the marker PH-RFP (a) PH-RFP-transfected Hec50 cellswere stimulatedwith the following ligands 10 nMestrogen (17120573E2) 10 120583M17120572-estrogen (17120572E2) or 10 nMestrogen in the presence of the PI3Kinhibitor LY294001 (10 120583M 20min pre-incubation 17120573E2 + LY) the EGFR inhibitor AG1478 (25 120583M 60min pre-incubation 17120573E2 + AG) orthe metalloproteinase inhibitor GM6001 (10120583M 30min pre-incubation 17120573E2 + GM) Unstimulated designates vehicle only (b) PH-RFP-transfected Hec50 cells were stimulated for 15min with the GPER-selective agonist G-1 or estrogen at the indicated concentrations in theabsence or presence of the GPER-selective antagonists G15 and G36 (cells were pretreated 15min with G15 or G36 prior to stimulation withE2 or G-1) or in the presence of 10 nM G-1 and the PI3K EGFR or metalloproteinase inhibitors as in (a)

and EGFR in estrogen-mediated PI3K activation in Hec50endometrial cancer cells

As Hec50 cells lack expression of the classical estrogenreceptor ER120572 but express GPER we next examined whetherthe activation of PI3K by estrogen might be mediated byGPER Using the GPER-selective agonist G-1 we observedthat like estrogen the PH-RFP reporter translocated tothe nucleus suggesting estrogen might be mediating itseffects via GPER (Figure 3(b)) In support of this the GPER-selective antagonists G15 and G36 not only prevented G-1-mediated activation of PI3K but also blocked estrogen-mediated PI3K activation (Figure 3(b)) G15 and G36 alonehad no effect As observed for estrogen-mediated activationof GPER PI3K activation in response to G-1 also requiresboth EGFR kinase and metalloproteinase activity as AG1478and GM6001 also blocked nuclear translocation of PH-RFPfollowing G-1 stimulation

To further demonstrate the requirement for GPER inPI3K activation by estrogen andG-1 beyond pharmacologicalinhibition we employed siRNA to knockdown expressionof GPER (Figure 4) In mock-transfected (no siRNA) andcontrol siRNA-transfected Hec50 cells both estrogen andG-1 stimulated nuclear localization of the PH-RFP reporterHowever in cells transfectedwithGPER-targeted siRNA nei-ther estrogen nor G-1 stimulation resulted in nuclear translo-cation of the PH-RFP reporter (Figure 4(a)) Knockdown

of GPER protein was confirmed by immunofluorescencestaining of mock control and GPER siRNA-transfected cells(Figure 4(b)) The use of both a pharmacological approach(G15 and G36) and siRNA to prevent activation of PI3K byestrogen as well as the ability of G-1 to activate PI3K stronglyindicates that GPER is the receptor mediating responsivenessto estrogen in Hec50 cells

33 Multiple Estrogen Mimetics Activate PI3K and ERK viaGPER To examine the effects of a number of therapeu-tic antiestrogens and other ligands on PI3K activation inER120572minus120573minus Hec50 cells we evaluated PH-RFP localizationin cells treated with 4-hydroxytamoxifen ICI182780 thebenzothiophene-based and recently FDA-approved SERMRaloxifene the phytoestrogen genistein and the widelyused ER120572- and ER120573-selective agonists propylpyrazoletriol(PPT) and diarylpropionitrile (DPN) (Figure 5(a)) There isevidence that ICI182780 [29] and tamoxifen [15 26 28]can act through GPER to stimulate rapid cellular signal-ing Specifically we have demonstrated tamoxifen-mediatedstimulation of PI3K in GPER-transfected COS7 cells andGPER+ SKBr3 breast cancer cells (both of which do notexpress ER120572 or for that matter ER120573) [28] PPT and DPNhave been used extensively as ldquoselectiverdquo agonists of ER120572 andER120573 respectively [63 64] Whereas PPT displays sim400-fold

Obstetrics and Gynecology International 7U

nstim

10nM

E2

10nM

G-1

Mock transfected GPER siRNAControl siRNA

(a)


GPE

R

(b)

Figure 4 GPERmediates PI3K activation in Hec50 cells (a) Hec50cells were transfected with no siRNA (mock transfected) controlsiRNA or siRNA targeting GPER (GPER siRNA) and the PH-RFP reporter Transfected Hec50 cells were stimulated with vehicleestrogen (10 nM E2) or G-1 (10 nM) (b) Hec50 cells transfectedwith no siRNA (mock transfected) and control siRNA or siRNAtargeting GPER (GPER siRNA) were stained for GPER (withcarboxy-terminal antibody) to demonstrate the specific knockdownof GPER in the GPER siRNA-treated cells

binding selectivity for ER120572 over ER120573 DPN exhibits onlysim70-fold selectivity for ER120573 over ER120572 [65ndash68] Of these allcompounds (at 100 nM) with the exception of DPN (evenat 10 120583M) stimulated the nuclear translocation of the PH-RFP reporter to the nucleus (Figure 5(a)) as observed withestrogen and G-1 (Figures 3 and 4) To confirm the activityof DPN we cotransfected Hec50 cells with ER120573-GFP andPH-RFP In cells expressing ER120573 DPN was indeed ableto stimulate PH-mRFP translocation at a concentration of100 nM demonstrating the ability ofHec50 cells to respond toDPN via ER120573 and furthermore demonstrating that withoutthe exogenous expression of ER120573 Hec50 cells do not expresssufficient ER120573 (if any) to respond to DPN

We and others have previously demonstrated that GPERis capable of activating ERK in addition to PI3K in multiplecancer and other cell lines including other endometrialcancer cell lines [22 27 42 69] To further examine andquantify the ligand specificity of GPER we determinedpERK levels in Hec50 cells stimulated with the ligands above

(Figures 5(b) and 5(c)) Although the only known estrogenreceptor expressed in Hec50 cells is GPER we confirmedthe contribution of GPER to ligand-induced ERK activa-tion using the GPER-selective antagonist G15 As for PI3Kactivation estrogen G-1 4-hydroxytamoxifen ICI182780and Raloxifene stimulated pERK sim5ndash8-fold Activation ofERK by each of these ligands was completely inhibited byG15 indicating an essential role for GPER in the responseto each ligand The phytoestrogen genistein also acted asan agonist of GPER-mediated ERK activation in Hec50cells and as observed with PI3K activation PPT (100 nM)but not DPN (at concentrations up to 10 120583M) was ableto induce ERK activation in a GPER-dependent (ie G15-sensitive) manner Finally to ensure that G15 did not inhibitERK activation downstream of the transactivated EGFR wedirectly stimulated Hec50 cells with EGF EGF was sim30more potent than the next most potent ligand estrogen butunlike the other agonists of ERK activation via GPER EGFstimulation was unaffected by G15 demonstrating that theinhibitory action of G15 on GPER activation is upstream ofEGFR

34 Expression of GPER in Xenografts of Endometrial CancerCell Lines and Human Endometrial Cancers Hec50 cellsare poorly differentiated endometrial cancer cells that wereoriginally isolated from a metastatic lesion in a patient withadvanced endometrial cancer who ultimately succumbed tothe cancer [70 71] The cells do not form glands in tissue cul-ture or in xenografts anddonot express either ER120572or PR [43]They do however exhibit the capacity to subdifferentiate intoa papillary serous phenotype when injected intraperitoneallyin mice [72] Thus Hec50 cells are an excellent model of typeII endometrial tumors [71] In contrast IshikawaH cells werederived from a patient with stage 2 moderately differentiatedendometrial adenocarcinoma who was treated with surgeryand chemotherapy and survived without recurrence Thesecells produce mucous contain vacuoles express both ER120572and PR and are thus an excellent model of type I endometrialcancer [71]

As demonstrated in Figure 1 by immunofluorescence andWestern blotting Hec50 cells express substantially (ge10-fold)more GPER than do H cells In xenografts H cells formendometrioid tumors whereas Hec50 can differentiate into aserous subtype [44] To assess whether GPER expression pat-terns are maintained in xenografts we performed immuno-histochemistry for GPER on xenograft tumors of bothH cellsand Hec50 cells from nude mice (Figure 6) Tumors derivedfrom H cells exhibited well-differentiated gland formation(Figures 6(a) and 6(b)) whereas in tumors derived fromHec50 cells cells were poorly or undifferentiated nuclei werepleomorphic and mitotic activity was abundant (Figures6(d) and 6(e)) In addition although GPER was detectedto varying extents in the tumors from H cells (Figures 6(a)and 6(b)) it was expressed at far greater levels in tumorsfrom Hec50 cells (Figures 6(d) and 6(e)) consistent withexpression observed in cultured monolayers (Figure 1) [28]

We have previously reported that in endometrial cancerhigh GPER expression is prognostic of poor survival [37]


4OH-tamoxifen ICI182780 Raloxifene

Genistein PPT DPN ER120573-GFP DPN

(a)

pERK

G15DMSO EGFDPNPPTGenICIRalTamG-1E2minus + minus + minus + minus + minus + minus + minus + minus + minus + minus +

(b)

lowast

lowast lowastlowast lowast

lowast

lowast

lowast

lowastlowast lowastlowast lowastlowast lowastlowast lowastlowastlowastlowast

lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)

Figure 5 GPER-mediated activation of PI3K and ERK in Hec50 cells by SERMs a SERD and an ER120572-selective agonist (a) PH-RFP-transfected Hec50 cells were stimulated for 15minutes with the following ligands 4-hydroxytamoxifen (100 nM 4OH-tamoxifen) ICI182780(100 nM) Raloxifene (100 nM) genistein (100 nM) PPT (100 nM) or DPN (10 120583M) Hec50 cells were also cotransfected with ER120573-GFP(shown in inset) and PH-RFP and stimulated with 100 nM DPN to verify the activity of DPN as an ER120573 agonist in Hec50 cells (ER120573-GFPDPN) (b) Hec50 cells were stimulated for 15min with vehicle (005 DMSO) estrogen (10 nM E2) G-1 (10 nM) 4-hydroxytamoxifen(100 nM Tam) Raloxifene (100 nM Ral) ICI182780 (100 nM ICI) genistein (100 nM Gen) PPT (100 nM) DPN (10 120583M) or EGF (1 nM)either in the presence or absence of 1120583MG15 (10min pretreatment with 005DMSO in samples without G15) (c) Band intensities of pERKwere normalized to total ERK and plotted with estrogen as 100 Data represent mean plusmn sem from three experiments lowast119875 lt 005 versusDMSO lowastlowast119875 lt 005 versus paired stimulus without G15

Moreover in carcinosarcoma subtypes advanced stage dis-ease wasmore frequently associated with high levels of GPERand ER120573 expression [73] To assess whether GPER expressionlevels and patterns are maintained in xenografts of pri-mary patient tumors xenografts from tumors preoperativelycharacterized as type I and type II tumors were generatedand immunostained for GPER using xenograft tissues and

parallel paraffin-embedded tissue from the original patienttumor (Figures 6(f)ndash6(l)) Case studies of our illustratedcases are relevant Patient 1 was originally diagnosed withsuperficially invasive (24) grade 1 ER120572+PR+ endometrioidadenocarcinoma (FIGO stage IB) and presented two yearslater at our institution with a 20 times 21 times 10 cm mass involvingthe omentum anterior abdominal wall and bowel whichwas


(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)

Figure 6 GPER expression in xenografts of Ishikawa H cells and Hec50 endometrial cancer cells as well as representative type I and typeII human primary tumors (a)ndash(e) Immunohistochemical staining of GPER in xenograft tumors of Ishikawa H cells ((a) 20x (b) 40x)demonstrating the areas most strongly positive for GPER and representative xenograft tumors of Hec50 cells ((d) 20x (e) 40x (c) negativecontrol (irrelevant primary antibody) 40x) ((f)ndash(h)) GPER staining of a recurrent adenocarcinoma with endometrioid features ((f) mousexenograft of patient tumor 40x (g) patient tumor 40x (h) patient tumor 20x demonstrating focal positivity (i) ER120572 and PR (inset) stainingof the same patient tumor) ((j)ndash(l)) GPER staining from a patient with Stage IA carcinosarcoma ((j) mouse xenograft of patient tumor 20xillustrating diffuse positive staining in both epithelial and stromal fractions (k) patient tumor 40x (l) patient tumor 20x demonstratingsimilarly strong GPER staining in both the epithelial and stromal compartments)

resected The patient refused postoperative chemotherapyand was instead treated with tamoxifen 40mg daily andmedroxyprogesterone acetate 200mg daily cycle day 16ndash30on an IRB-approved institutional trial With 6 months ofhormonal therapy the patient has remained radiographicallyfree of disease (duration of followup 45 years) Immunos-taining of the recurrent tumor (the primary was unavailablefor comparison) was scored as PR (3+ 10 of viable epithelialcells) ER120572 (3+ 100 of viable cells) and GPER (3+ less than10 of viable cells Figures 6(f)ndash6(i))

Patient 2 was diagnosed with grade 3 adenocarcinomabased upon endometrial biopsy but at surgery was found tohave carcinosarcoma FIGO stage IA with high-grade sar-comatous features Following weekly cisplatin chemotherapyand radiation therapy this patient has remained disease-free 4years after therapy As carcinosarcoma tumors are consideredbiphasic (ie defined by having malignant epithelial andstromal compartments) [37 74] we evaluated both epithelialand stromal cell staining for GPER expression and observedthat both components were strongly positive for GPER


800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)

Figure 7 Inhibition of Hec50 tumor growth by GPER antagonist Hec50 cell tumors were initiated in ovariectomized athymic mice andtreated with a slow release pellet containing either no added compound (sham) estrogen (E2) G-1 or with two pellets (one of estrogen andone of G36 E2 + G36) (a) Tumor volume was measured with calipers over a 24-day period (b) Upon sacrifice tumors were dissected andtumor mass was determined Data represent mean plusmn sem from 6ndash8 mice lowast119875 lt 005 versus sham 119875 lt 005 versus E2 alone Note that thetumor size for E2 + G36 was not significantly different from the sham

These results demonstrate not only that Hec50 cells maintainhigh levels of GPER expression as xenografted tumors butalso that Hec50 cells mimic the levels of GPER expressionobserved in high grade endometrial tumors as detected inparaffin-embedded patient samples and xenografted patienttumors suggesting that primary xenografts of endometrialcancers (even complex tumors with biphasic characteristics)may represent an excellent model to test the therapeuticefficacy of GPER-targeted therapies

Interestingly within our endometrial cancer repositorywe identified six patients who received tamoxifen-basedtherapy for recurrent endometrial cancer and of these onlyone patient (case 1 above with low GPER expression) expe-rienced a complete response in contrast all nonrespondersdisplayed increased GPER expression (defined as expressionabove the mean [37]) by immunohistochemistry of paraffin-embedded tumor samples (Fisherrsquos 119875 = 003) suggestingthat a lack of or low GPER expression may be a predictor oftamoxifen responsiveness in endometrial cancer

35 GPER Antagonist Inhibits Endometrial Tumor GrowthType I endometrial cancer can result from excess andorunopposed estrogen use and typically progresses from hyper-plasia to atypical hyperplasia and finally to carcinomaType I tumors commonly express ER and PR [75] and aregenerally responsive to hormone treatment with therapeuticefficacy positively correlating with the level of receptorexpression [2 75 76] However type II endometrial can-cers are believed to develop through molecular pathwaysinvolving p53 mutations which more closely resemble high-grade serous ovarian tumors in molecular alterations andmorphology and commonly do not express either ER or PR[2 44] suggesting that estrogen (or its inhibition) should

have no effect on the growth of such tumors To examine theestrogen dependence of a type II endometrial cancer in vivowe generatedHec50 cell xenografts inmice To determine theeffect of estrogen the tumorswere initiated in ovariectomizedmice and estrogen was restored using slow release pellets(Figure 7) Surprisingly we found very little tumor growth inovariectomized mice (compared to our experience in ovaryintact mice data not shown) However upon supplementa-tion with estrogen tumor growth was increased about 10-fold (compared to ovariectomizedsham-treated mice) Todetermine whether GPER was responsible for this estrogen-mediated enhancement of tumor growth we also treatedovariectomized mice with slow release pellets containingG-1 The tumors in these mice were about 5-fold largerthan those in sham-treated mice suggesting that GPER wasindeed capable of stimulating tumor growth in vivo Finallyto further test whether the estrogen-stimulated tumor growthwasmediated by GPER we cotreated estrogen-supplementedmice with our recently identified GPER-selective antagonistG36 [42] Estrogen-stimulated tumor growth was reduced byG36 to that of the sham treatment (ie estrogen deprivedstate) demonstrating a critical role for GPER in the estrogen-mediated response of ERminus type II endometrial cancer growth

4 Discussion

Endometrial cancers like most cancers consist of multi-ple distinguishable tumor types At a minimum endome-trial cancers have been separated into type I and type IIendometrial cancers based on morphology and molecularphenotypesgenotypes [2] Recent genomic studies fromTheCancer Genome Atlas are shedding more light on the extentand range of mutations associated with this cancer defining


4 categories of endometrial cancer [77] Whereas survivalis high among women with type I endometrial cancers theopposite is true for type II cancers which express high levelsof GPER [37] We have therefore focused our investigationson a model of type II endometrial cancer in this studyparticularly due to the fact that these cancers are typicallydescribed as estrogen unresponsive due to their lack of ERexpression In this work we have demonstrated that Hec50cells typical of type II endometrial cancer cells that do notexpress the classical ER120572 do express GPER which makesthem responsive to estrogen in terms of rapid cellular sig-naling (PI3K and ERK) Furthermore as chronic tamoxifenuse (ie for breast cancer) results in an increased incidenceof endometrial cancer we demonstrated that in endometrialcancer cells GPER mediates cellular signaling in response totwo SERMs (tamoxifen and Raloxifene) as well as a SERD(ICI182780) revealing a possible additional mechanism forthe increased risk of endometrial cancer with tamoxifen useWe also demonstrated that the ER120572-selective (ie selectiveversus ER120573) agonist PPT activates GPER raising questionsabout the conclusions drawn from its use in defining exclusiveER120572 function in a multitude of biological systems Finally weestablished that GPER is highly expressed in type II tumorsthat Hec50 xenograft tumors display a strong dependenceon estrogen in vivo and that this occurs through GPERthe inhibition of which blocks estrogen-stimulated tumorgrowth

Although GPCRs are traditionally thought of as cellsurface receptors mediating transmembrane signaling ofmembrane-impermeable ligands (eg ionic small moleculespeptides and proteins) we originally described GPER local-ization as being predominantly intracellular [28] Those andour continuing results have localized GPER primarily tointracellular membrane compartments (endoplasmic reticu-lum and Golgi membranes) even using immunohistochem-istry of tumor and other tissues [34 37 53 54] In Hec50cells we again observed a strong intracellular localizationTo examine whether this localization was also evident inprimary cells (as opposed to cancer cell lines) we isolatedand stained primary murine uterine epithelial cells As withHec50 cells the staining pattern of GPER appeared intracel-lular but with a more punctate morphology Consistent withan intracellular localization no staining was observed usingan antibody targeting the amino terminus of GPER undernonpermeabilizing conditions In a fraction of cells stainingof the nuclear membrane was evident As the endoplasmicreticulum is continuous with the nuclear membrane sucha pattern would not be unexpected However since nuclearmembrane staining is not present in all cells and cell typesGPER localization to the nuclear and other membranesmay be actively regulated In fact others have reportedplasma membrane localization to varying extents in diversecell types [56ndash58] Whether the localization and traffickingare dynamically regulated within a cell or simply differentin distinct cell types remains to be determined The lackof extensive plasma membrane localization under steady-state conditions suggests either that GPER traffics poorly tothe plasma membrane (being retained in the endoplasmicreticulum and Golgi apparatus) or that if trafficked to the

plasma membrane it is rapidly internalized Evidence forboth of these mechanisms has been recently presented withFilardo et al revealing constitutive internalization of GPERto a trans-Golgi compartment [78] and with Lenhart et alsuggesting a role for receptor activity-modifying protein 3(RAMP3) in the trafficking of GPER to the cell surface[79] With the identification of an increasing number ofGPCRs being expressed intracellularly [80 81] particularlythose GPCRs for membrane-permeable ligands (such aslipids and steroids) [82] as well as the recognized activitiesof internalized GPCRs [83] the cellular mechanisms andfunctional consequences of regulating GPER localizationremain to be elucidated With the recent characterizationof differential subcellular activation of calcium stores byGPR55 depending on its site of activation in cardiomyocytes(plasmalemmal versus intracellular) [82] the subcellular siteof GPER expressionactivation may similarly play an impor-tant role in regulating its downstream signaling activityparticularly given the broad scope of GPER expression andfunction throughout the body [16 69 84ndash87]

Whereas the classical estrogen receptors (ER120572 and ER120573)are traditionally thought to mediate primarily genomicresponses [88] GPER has become recognized as an estrogenreceptor thatmediates rapid cellular signaling [89]Neverthe-less there is also substantial evidence that ER120572 can mediaterapid signaling [90] and that GPER mediates transcriptionalregulation [91] Whereas ER120572 regulates transcription in partthrough direct binding to estrogen response elements inDNA [88] GPER presumably regulates transcription indi-rectly through kinase cascades [91]While the transcriptionalmechanisms of each receptor are distinct there exist manypossibilities for overlap and interactions of signaling activities(both synergistic and inhibitory) that remain largely unde-fined The end results of GPER activation are also likely to bedifferent in cells that also express ER120572 (or ER120573 for thatmatter)compared to cells that express only GPER For example therole of GPER in ER120572+ breast cancer cells that are ldquoaddictedrdquoto ER-mediated signalinggene expression for growth andthus sensitive to antiestrogenhormone therapy may be verydifferent compared to ERminustriple negative cells The samemay be true of endometrial cancers with ER120572+ type I tumorsbehaving very differently compared to ERminus type II tumorsparticularly with respect to estrogen signaling Furthermorethe effects of additional growth promoting factorspathwaysmay obscure or minimize the effects of estrogen throughGPER or ERs for that matter

In this study we demonstrated that an endometrialcell line representative of ERminus type II tumors maintainsthe ability to signal in response to estrogen via GPERIn Hec50 cells estrogen signaling via GPER results ina metalloproteinaseEGFR-dependent activation of down-stream kinase pathways including PI3K and ERK bothimportant players in cancer cell survival and growth Inter-estingly the plasma membrane localization of PIP3 underunstimulated conditions suggests a level of constitutive PI3Kactivity in Hec50 cells The ability of an EGFR inhibitor toreduce this further suggests that one or more members of theerbB family are involved through EGFR activation [92] It isimportant to note that the effect of estrogen and G-1 might


not be easily detectable by measuring total cellular pAktlevels for example by Western blot due to the constitutiveactivation of PI3K Translocation of the PH-RFP reporterhowever allows selective detection of GPER-mediated PI3Kactivation in the absence of global changes in the total cellularlevel of PI3K activity Such locationcompartment-specificsignaling activity may play an important role in the overalleffect of any activated pathway

The pharmacopeia of estrogen receptor ligands hasbeen developed in the absence of consideration for theirinteraction with GPER For example the development ofSERMs and SERDs was largely based on the developmentof compounds with high binding affinity to ER120572 and theeffects of these compounds on transcriptional regulation(through estrogen response elements) and overall tissue-specific activities (breast cancer cell growth versus uterineeffects such as imbibition) Furthermore the search for ER120572versus ER120573 (and vice versa) selectivity [68 93 94] has alsobeen carried without consideration of GPER function Inmany cases ligands or therapeutic agents were developedlong before GPER was even identified tamoxifen dating backto the 1970s Inmany studies these agents exhibit unexpectedactivities often stimulating rapid cellular or physiologicalresponses similar to estrogen without the expected inhibitoryeffects [95ndash100] Although the disparate activities of SERMshave been attributed to tissue differences in the expressionof ER coregulators [101] the agonistic effects of SERDs areperhaps less easily explained by such mechanisms With pre-vious reports of tamoxifen [28] and ICI182780 [29] agonismthrough GPER and now with our current demonstration ofthe activity of Raloxifene onGPER themechanisms of actionof these compounds are greatly complicated For examplein addition to ER120572 what role does GPER play in hormoneresistance in breast cancer [30 31 102] and the increasedincidence of endometrial pathology and cancer in womentaking tamoxifen [32] The agonism of these compounds onGPER may well play an important role

With the identification of ER120573 in 1996 [103] it soonbecame clear that selective ligands would be a powerful toolin the characterization of the functions of the individualestrogen receptors as well as being therapeutically promisingBecause the ligand binding pockets of ER120572 and ER120573 arealmost identical achieving this goal has been challenging[104] Today the most widely used ldquoselectiverdquo agonists forER120572 and ER120573 are propylpyrazoletriol (PPT) and diarylpro-pionitrile (DPN) respectively [63 64] PPT displays sim400-fold binding selectivity for ER120572 over ER120573 whereas DPNexhibits only sim70-fold selectivity for ER120573 over ER120572 [65ndash68]In our current study we found that PPT (at 100 nM witha weak response at 10 nM unpublished observation) alsoacts as an agonist for GPER The three estrogen receptorselective compounds PPT DPN and G-1 have recently beenused to evaluate the estrogen receptor involved in a numberof physiological or cellular responses [105ndash107] Prior to theidentification of G-1 only PPT and DPN could be used toldquodistinguishrdquo between ER120572 and ER120573 A growing number ofreports are concluding for example that both ER120572 andGPERmediate responses based on the activity of both PPT andG-1 [108 109] however based on our current results these

responses could be mediated solely by GPER if other mea-sures of receptor involvement are not also employed ClearlyPPT is a potent agonist of ER120572 with enhanced selectivityversus GPER but without additional approaches (siRNAor selective antagonists such as G15 and G36) conclusionsmust be drawn with care and it is possible that many ofthe interpretations in the related literature (particularly usinghigh PPT concentrations eg ge10 nM) should be reevaluatedwith these considerations in mind

GPER stimulation has been demonstrated to increasecell proliferation in a broad array of cell lines includingsome endometrial cancer cell lines suggesting a potentialimportance in one or more aspects of carcinogenesis [1520 24 26 27] In fact in a very recent study knockingdown GPER expression in Hec1A endometrial cancer cells(as well as Ishikawa H cells) resulted in a reduction oftumor growth in athymic mice [20] This result is consis-tent with the many studies that suggest GPER expressioncorrelates with poor survival or indicators of poor out-come in endometrial ovarian and other cancers [33ndash37]In particular GPER expression represents a mechanism bywhich ostensibly estrogen-unresponsive tumors (often statedas such based on the lack of ER120572 expression alone) canmaintain estrogen responsiveness The significance of thisestrogen responsiveness will clearly depend on the type ofcancer and the other mechanismsmutations involved in aspecific cancer Although it has been amply demonstratedthat in breast cancer antiestrogens targeting ER120572 are highlyeffective the role of estrogen in gynecological cancers isthought to be of less importance due to the lack of clear effi-cacy of antiestrogens [110] Nevertheless the endometriumlike the breast is highly estrogen responsive in terms ofproliferation and elevated tumor estrogen levels have beenreported not only in ER120572+ type I but also in ER120572minus typeII endometrial tumors [111] Thus if GPER expression andfunction play an important role particularly in ER120572minus typeII endometrial cancers then treatment with SERMs andSERDs functioning as GPER agonists would be highlycontraindicated

To examine the role of GPER in endometrial tumorgrowth we sought to establish and investigate xenografttumors in mice We first examined whether the highlydisparate levels of GPER expression in H cells and Hec50cells observed in tissue cultureweremaintained as xenograftsOverall xenograft tumors of each cell type displayed theexpected histological properties with respect to tumor mor-phology and GPER expression with Hec50 cell tumorsexpressing substantially higher levels compared to H celltumors Importantly primary xenografts displayed similarmorphologies and GPER expression levels and patterns tothose assessed directly from the patient samplesThis suggeststhat the xenograft model may be a useful adjunct in whichto test the therapeutic efficacy of GPER-selective compounds(particularly GPER-selective antagonists) Of particular rele-vance in the clinical setting we found that low expressionof GPER was associated with the only observed responseto tamoxifen and since this patient was strongly ER+PR+we surmise this effect may have been mediated by classicalsteroid receptor pathways [112]


To determine if estrogen itself and GPER in particu-lar contributes to tumor growth of ER120572minus Hec50 cells weestablished subcutaneous Hec50 tumors in ovariectomizedathymic mice Tumors remained small in untreated micebut were about 10-fold larger in mice treated with slowrelease estrogen pellets indicating that the lack of ER120572 didnot prevent responsiveness to estrogen Interestingly wewere not able to demonstrate estrogen-stimulated growthin monolayer culture of Hec50 cells suggesting estrogen-responsiveness may be a result of the tumor environ-ment (unpublished results) highlighting the importance oftumorigenesis studies in vivo To test whether GPER canmediate this estrogen responsiveness we treated mice withslow release pellets containing G-1 These tumors were abouthalf the size of the estrogen-treated mice suggesting thatGPER could mediate the effects of estrogen The reducedtumor size of G-1-treated mice compared to estrogen-treatedmice could be due to differential sensitivity to the two ligands(eg EC

50values) or differential distribution metabolism or

excretion of the two compounds Finally to investigate moredirectly the role of GPER in the estrogen responsivenesswe treated mice with both estrogen and the GPER-selectiveantagonist G36 The tumors in these mice were restored tothe size of the tumors in the sham-treated miceThese resultsindicate not only that the observed estrogen dependence ofHec50 cell tumors is due to GPER but also that pharmaco-logical inhibition of GPER could represent an important newtherapy for women particularly those with aggressive type IIendometrial cancer

5 Conclusions

In this paper we have demonstrated that GPER plays animportant role in the estrogen-mediated signaling of a rep-resentative type II endometrial cancer cell line In additionwe demonstrate for the first time that the SERMRaloxifene isan agonist for GPER with potentially important clinical ram-ifications for its FDA-approved chronic use for osteoporosisThe ability of the widely used ldquoER120572-selectiverdquo agonist PPTto activate GPER suggests that a large body of literature mayhave to be more carefully interpreted with respect to definingthe roles of individual estrogen receptors Finally we havedemonstrated that GPER-selective antagonists may representimportant new therapeutic agents for endometrial and othercancers pathologically defined as estrogen unresponsive dueto their lack of ER120572 expression

Authorsrsquo Contribution

Whitney K Petrie andMegan K Dennis contributed equally

Acknowledgments

This work was funded by NIH Grants CA118743 andCA163890 to Eric R Prossnitz and CA127731 to EricR Prossnitz and Jeffrey B Arterburn Images in thispaper were generated in the University of New Mexico ampCancer Center Fluorescence Microscopy Shared Resource

funded as detailed on httphscunmeducrtcmicroscopyacknowledgementshtml

References

[1] J L Hecht and G L Mutter ldquoMolecular and pathologic aspectsof endometrial carcinogenesisrdquo Journal of Clinical Oncologyvol 24 no 29 pp 4783ndash4791 2006

[2] L S G Ulrich ldquoEndometrial cancer types prognosis femalehormones and antihormonesrdquo Climacteric vol 14 no 4 pp418ndash425 2011

[3] M G del Carmen M Birrer and J O Schorge ldquoUterinepapillary serous cancer a review of the literaturerdquo GynecologicOncology vol 127 no 3 pp 651ndash661 2012

[4] MKMcConechy J DingMC Cheang et al ldquoUse ofmutationprofiles to refine the classification of endometrial carcinomasrdquoThe Journal of Pathology vol 228 no 1 pp 20ndash30 2012

[5] AGarrett andMAQuinn ldquoHormonal therapies and gynaeco-logical cancersrdquo Best Practice and Research in Clinical Obstetricsand Gynaecology vol 22 no 2 pp 407ndash421 2008

[6] T Thigpen M F Brady H D Homesley J T Soper and JBell ldquoTamoxifen in the treatment of advanced or recurrentendometrial carcinoma a Gynecologic Oncology Group studyrdquoJournal of Clinical Oncology vol 19 no 2 pp 364ndash367 2001

[7] J V Pinkerton and S R Goldstein ldquoEndometrial safety a keyhurdle for selective estrogen receptor modulators in develop-mentrdquoMenopause vol 17 no 3 pp 642ndash653 2010

[8] A E Bland B Calingaert A A Secord et al ldquoRelationshipbetween tamoxifen use and high risk endometrial cancerhistologic typesrdquo Gynecologic Oncology vol 112 no 1 pp 150ndash154 2009

[9] A I Tergas R Buell-Gutbrod K Gwin et al ldquoClinico-pathologic comparison of type II endometrial cancers based ontamoxifen exposurerdquo Gynecologic Oncology vol 127 no 2 pp316ndash320 2012

[10] H Zhang TMcElrathWTong and JW Pollard ldquoThemolecu-lar basis of tamoxifen induction of mouse uterine epithelial cellproliferationrdquo Journal of Endocrinology vol 184 no 1 pp 129ndash140 2005

[11] Y Shang ldquoMolecular mechanisms of oestrogen and SERMs inendometrial carcinogenesisrdquo Nature Reviews Cancer vol 6 no5 pp 360ndash368 2006

[12] C L Smith and B W OrsquoMalley ldquoCoregulator function a keyto understanding tissue specificity of selective receptor modu-latorsrdquo Endocrine Reviews vol 25 no 1 pp 45ndash71 2004

[13] Y Shang andM Brown ldquoMolecular determinants for the tissuespecificity of SERMsrdquo Science vol 295 no 5564 pp 2465ndash24682002

[14] V Gigoux and D Fourmy ldquoActing on hormone receptors withminimal side effect on cell proliferation a timely challengeillustrated with GLP-1R and GPERrdquo Frontiers in Endocrinologyvol 4 p 50 2013

[15] A Vivacqua D Bonofiglio A G Recchia et al ldquoTheG protein-coupled receptor GPR30 mediates the proliferative effectsinduced by 17120573-estradiol and hydroxytamoxifen in endometrialcancer cellsrdquo Molecular Endocrinology vol 20 no 3 pp 631ndash646 2006

[16] E R Prossnitz and M Barton ldquoThe G-protein-coupled estro-gen receptor GPER in health and diseaserdquo Nature ReviewsEndocrinology vol 7 no 12 pp 715ndash726 2011


[17] E R Prossnitz J B Arterburn H O Smith T I Oprea LA Sklar and H J Hathaway ldquoEstrogen signaling through thetransmembrane G protein-coupled receptor GPR30rdquo AnnualReview of Physiology vol 70 pp 165ndash190 2008

[18] L Albanito A Madeo R Lappano et al ldquoG protein-coupledreceptor 30 (GPR30) mediates gene expression changes andgrowth response to 17120573-estradiol and selective GPR30 ligandG-1 in ovarian cancer cellsrdquo Cancer Research vol 67 no 4 pp1859ndash1866 2007

[19] A Vivacqua D Bonofiglio L Albanito et al ldquo17120573-Estradiolgenistein and 4-hydroxytamoxifen induce the proliferation ofthyroid cancer cells through the G protein-coupled receptorGPR30rdquoMolecular Pharmacology vol 70 no 4 pp 1414ndash14232006

[20] X Ge R Guo Y Qiao et al ldquoThe G protein-coupled receptorGPR30 mediates the nontranscriptional effect of estrogen onthe activation of PI3KAkt pathway in endometrial cancer cellsrdquoInternational Journal of Gynecological Cancer vol 23 no 1 pp52ndash59 2013

[21] Y Wei Z Zhang H Liao et al ldquoNuclear estrogen receptor-mediated Notch signaling and GPR30-mediated PI3KAKTsignaling in the regulation of endometrial cancer cell prolifera-tionrdquo Oncology Reports vol 27 no 2 pp 504ndash510 2012

[22] A Vivacqua E Romeo P De Marco E M De FrancescoS Abonante and M Maggiolini ldquoGPER mediates the Egr-1expression induced by 17120573-estradiol and 4-hydroxitamoxifen inbreast and endometrial cancer cellsrdquoBreast Cancer Research andTreatment vol 133 no 3 pp 1025ndash1035 2012

[23] Y Y He G Q Du B Cai et al ldquoEstrogenic transmembranereceptor of GPR30 mediates invasion and carcinogenesis byendometrial cancer cell line RL95-2rdquo Journal of Cancer Researchand Clinical Oncology vol 138 no 5 pp 775ndash783 2012

[24] G-Q Du L Zhou X-Y Chen X-P Wan and Y-Y He ldquoThe Gprotein-coupled receptor GPR30 mediates the proliferative andinvasive effects induced by hydroxytamoxifen in endometrialcancer cellsrdquo Biochemical and Biophysical Research Communi-cations vol 420 no 2 pp 343ndash349 2012

[25] N Filigheddu S Sampietro F Chianale et al ldquoDiacylglycerolkinase 120572mediatses 17-120573-estradiol-induced proliferation motil-ity and anchorage-independent growth of Hec-1A endometrialcancer cell line through the G protein-coupled estrogen recep-tor GPR30rdquo Cellular Signalling vol 23 no 12 pp 1988ndash19962011

[26] B C Lin M Suzawa R D Blind et al ldquoStimulating the GPR30estrogen receptor with a novel tamoxifen analogue activates SF-1 and promotes endometrial cell proliferationrdquoCancer Researchvol 69 no 13 pp 5415ndash5423 2009

[27] Y-Y He B Cai Y-X Yang X-L Liu and X-P Wan ldquoEstro-genic G protein-coupled receptor 30 signaling is involved inregulation of endometrial carcinoma by promoting prolifer-ation invasion potential and interleukin-6 secretion via theMEKERK mitogen-activated protein kinase pathwayrdquo CancerScience vol 100 no 6 pp 1051ndash1061 2009

[28] CM Revankar D F Cimino L A Sklar J B Arterburn and ER Prossnitz ldquoA transmembrane intracellular estrogen receptormediates rapid cell signalingrdquo Science vol 307 no 5715 pp1625ndash1630 2005

[29] E J Filardo J A Quinn K I Bland and J FrackeltonAR ldquoEstrogen-induced activation of Erk-1 and Erk-2 requiresthe G protein-coupled receptor homolog GPR30 and occursvia trans-activation of the epidermal growth factor receptor

through release of HB-EGFrdquo Molecular Endocrinology vol 14no 10 pp 1649ndash1660 2000

[30] A Ignatov T Ignatov A Roessner S D Costa and T KalinskildquoRole of GPR30 in the mechanisms of tamoxifen resistancein breast cancer MCF-7 cellsrdquo Breast Cancer Research andTreatment vol 123 no 1 pp 87ndash96 2010

[31] A Ignatov T Ignatov C Weienborn et al ldquoG-protein-coupledestrogen receptor GPR30 and tamoxifen resistance in breastcancerrdquo Breast Cancer Research and Treatment vol 128 no 2pp 457ndash466 2011

[32] T Ignatov H Eggemann A Semczuk et al ldquoRole of GPR30in endometrial pathology after tamoxifen for breast cancerrdquoAmerican Journal of Obstetrics and Gynecology vol 203 no 6pp 595e9ndash595e16 2010

[33] E J Filardo C T Graeber J A Quinn et al ldquoDistributionof GPR30 a seven membrane-spanning estrogen receptor inprimary breast cancer and its associationwith clinicopathologicdeterminants of tumor progressionrdquo Clinical Cancer Researchvol 12 no 21 pp 6359ndash6366 2006

[34] H O Smith H Arias-Pulido D Y Kuo et al ldquoGPR30 predictspoor survival for ovarian cancerrdquoGynecologic Oncology vol 114no 3 pp 465ndash471 2009

[35] V R Jala B N Radde B Haribabu and C M KlingeldquoEnhanced expression of G-protein coupled estrogen receptor(GPERGPR30) in lung cancerrdquo BMC Cancer vol 12 article624 2012

[36] J P Glass G Parasher H Arias-Pulido R Donohue L A Cer-illi and E R Prossnitz ldquoMesothelin andGPR30 staining amonga spectrum of pancreatic epithelial neoplasmsrdquo InternationalJournal of Surgical Pathology vol 19 no 5 pp 588ndash596 2011

[37] H O Smith K K Leslie M Singh et al ldquoGPR30 a novel indi-cator of poor survival for endometrial carcinomardquo AmericanJournal of Obstetrics and Gynecology vol 196 no 4 pp 386e1ndash386e11 2007

[38] Z Kolkova V Casslen E Henic et al ldquoThe G protein-coupledestrogen receptor 1 (GPERGPR30) does not predict survival inpatients with ovarian cancerrdquo Journal of Ovarian Research vol5 article 9 2012

[39] T Ignatov C Weissenborn A Poehlmann et al ldquoGPER-1 expression decreases during breast cancer tumorigenesisrdquoCancer Investigation vol 31 no 5 pp 309ndash315 2013

[40] C G Bologa C M Revankar S M Young et al ldquoVirtualand biomolecular screening converge on a selective agonist forGPR30rdquo Nature Chemical Biology vol 2 no 4 pp 207ndash2122006

[41] M K Dennis R Burai C Ramesh et al ldquoIn vivo effects of aGPR30 antagonistrdquo Nature Chemical Biology vol 5 no 6 pp421ndash427 2009

[42] M K Dennis A S Field R Burai et al ldquoIdentification ofa GPERGPR30 antagonist with improved estrogen receptorcounterselectivityrdquo Journal of Steroid Biochemistry and Molec-ular Biology vol 127 no 3ndash5 pp 358ndash366 2011

[43] N S Kumar J Richer G Owen E Litman K B Horwitzand K K Leslie ldquoSelective down-regulation of progesteronereceptor isoform B in poorly differentiated human endometrialcancer cells implications for unopposed estrogen actionrdquo Can-cer Research vol 58 no 9 pp 1860ndash1865 1998

[44] L Albitar G Pickett M Morgan S Davies and K K LeslieldquoModels representing type I and type II human endometrialcancers ishikawa H and Hec50co cellsrdquo Gynecologic Oncologyvol 106 no 1 pp 52ndash64 2007


[45] T InabaWGWiest R C Strickler and JMori ldquoAugmentationof the response of mouse uterine epithelial cells to estradiol byuterine stromardquo Endocrinology vol 123 no 3 pp 1253ndash12581988

[46] T Balla and P Varnai ldquoVisualizing cellular phosphoinositidepools with GFP-fused protein-modulesrdquo Sciencersquos STKE vol2002 no 125 article pl3 2002

[47] R E Campbell O Tour A E Palmer et al ldquoA monomericred fluorescent proteinrdquo Proceedings of the National Academyof Sciences of the United States of America vol 99 no 12 pp7877ndash7882 2002

[48] D Dai A M Holmes T Nguyen et al ldquoA potential synergisticanticancer effect of paclitaxel and amifostine on endometrialcancerrdquo Cancer Research vol 65 no 20 pp 9517ndash9524 2005

[49] W Luo F Wu R Elmaoued et al ldquoAmifostine enhancementof the anti-cancer effects of paclitaxel in endometrial cancer isTP53-dependentrdquo International Journal of Oncology vol 37 no5 pp 1187ndash1194 2010

[50] A Madeo and M Maggiolini ldquoNuclear alternate estrogenreceptor GPR30 mediates 17120573-estradiol - Induced gene expres-sion and migration in breast cancer - Associated fibroblastsrdquoCancer Research vol 70 no 14 pp 6036ndash6046 2010

[51] S Chakrabarti and S T Davidge ldquoG-protein coupled receptor30 (GPR30) a novel regulator of endothelial inflammationrdquoPLoS One vol 7 no 12 Article ID e52357 2012

[52] C Otto B Rohde-Schulz G Schwarz et al ldquoG protein-coupledreceptor 30 localizes to the endoplasmic reticulum and is notactivated by estradiolrdquo Endocrinology vol 149 no 10 pp 4846ndash4856 2008

[53] H Sakamoto K-I Matsuda K Hosokawa et al ldquoExpression ofG protein-coupled receptor-30 a G protein-coupledmembraneestrogen receptor in oxytocin neurons of the rat paraventricularand supraoptic nucleirdquo Endocrinology vol 148 no 12 pp 5842ndash5850 2007

[54] K Matsuda H Sakamoto H Mori et al ldquoExpression andintracellular distribution of the G protein-coupled receptor 30in rat hippocampal formationrdquo Neuroscience Letters vol 441no 1 pp 94ndash99 2008

[55] C S Liverman J W Brown R Sandhir K E McCarson andN E J Berman ldquoRole of the oestrogen receptors GPR30 andER120572 in peripheral sensitization relevance to trigeminal paindisorders in womenrdquo Cephalalgia vol 29 no 7 pp 729ndash7412009

[56] C Sanden S Broselid L Cornmark et al ldquoG protein-coupledestrogen receptor 1G protein-coupled receptor 30 localizes inthe plasmamembrane and traffics intracellularly on cytokeratinintermediate filamentsrdquoMolecular Pharmacology vol 79 no 3pp 400ndash410 2011

[57] T Funakoshi A Yanai K Shinoda M M Kawano andY Mizukami ldquoG protein-coupled receptor 30 is an estrogenreceptor in the plasmamembranerdquo Biochemical and BiophysicalResearch Communications vol 346 no 3 pp 904ndash910 2006

[58] E Filardo J Quinn Y Pang et al ldquoActivation of the novelestrogen receptor G protein-coupled receptor 30 (GPR30) atthe plasma membranerdquo Endocrinology vol 148 no 7 pp 3236ndash3245 2007

[59] S-B Cheng J A Quinn C T Graeber and E J FilardoldquoDown-modulation of the G-protein-coupled estrogen recep-tor GPER from the cell surface occurs via a trans-golgi-proteasome pathwayrdquo The Journal of Biological Chemistry vol286 no 25 pp 22441ndash22455 2011

[60] S-B Cheng C T Graeber J A Quinn and E J Filardo ldquoRet-rograde transport of the transmembrane estrogen receptor G-protein-coupled-receptor-30 (GPR30GPER) from the plasmamembrane towards the nucleusrdquo Steroids vol 76 no 9 pp 892ndash896 2011

[61] H Banfic D Visnjic N Mise et al ldquoEpidermal growth factorstimulates translocation of the class II phosphoinositide 3-kinase PI3K-C2120573 to the nucleusrdquo Biochemical Journal vol 422no 1 pp 53ndash60 2009

[62] A M Martelli L Cocco S Capitani S Miscia S Papa and FA Manzoli ldquoNuclear phosphatidylinositol 345-trisphosphatephosphatidylinositol 3-kinase Akt and PTen emerging keyregulators of anti-apoptotic signaling and carcinogenesisrdquoEuro-pean Journal of Histochemistry vol 51 pp 125ndash131 2007

[63] W R Harrington S Sheng D H Barnett L N Petz J AKatzenellenbogen and B S Katzenellenbogen ldquoActivities ofestrogen receptor alpha- and beta-selective ligands at diverseestrogen responsive gene sites mediating transactivation ortransrepressionrdquo Molecular and Cellular Endocrinology vol206 no 1-2 pp 13ndash22 2003

[64] J Frasor D H Barnett J M Danes R Hess A F Parlow andB S Katzenellenbogen ldquoResponse-specific and ligand dose-dependent modulation of estrogen receptor (ER) 120572 activity byER120573 in the uterusrdquo Endocrinology vol 144 no 7 pp 3159ndash31662003

[65] J Sun J Baudry J A Katzenellenbogen and B S Katzenel-lenbogen ldquoMolecular basis for the subtype discrimination ofthe estrogen receptor-120573-selective ligand diarylpropionitrilerdquoMolecular Endocrinology vol 17 no 2 pp 247ndash258 2003

[66] M J Meyers J Sun K E Carlson G A Marriner BS Katzenellenbogen and J A Katzenellenbogen ldquoEstrogenreceptor-120573 potency-selective ligands structure-activity rela-tionship studies of diarylpropionitriles and their acetylene andpolar analoguesrdquo Journal ofMedicinal Chemistry vol 44 no 24pp 4230ndash4251 2001

[67] S R Stauffer C J Coletta R Tedesco et al ldquoPyrazole ligandsstructuremdashaffinityactivity relationships and estrogen receptor-120572-selective agonistsrdquo Journal ofMedicinal Chemistry vol 43 no26 pp 4934ndash4947 2000

[68] J Sun M J Meyers B E Fink R Rajendran J A Katzenellen-bogen and B S Katzenellenbogen ldquoNovel ligands that functionas selective estrogens or antiestrogens for estrogen receptor-120572or estrogen receptor-120573rdquo Endocrinology vol 140 no 2 pp 800ndash804 1999

[69] G Sharma and E R Prossnitz ldquoMechanisms of estradiol-induced insulin secretion by the G protein-coupled estrogenreceptor GPR30GPER in pancreatic 120573-cellsrdquo Endocrinologyvol 152 no 8 pp 3030ndash3039 2011

[70] M Suzuki H Kuramoto andMHamano ldquoEffects of oestradioland progesterone on the alkaline phosphatase activity of ahuman endometrial cancer cell-linerdquoActa Endocrinologica vol93 no 1 pp 108ndash113 1980

[71] L Albitar L L Laidler R Abdallah and K K Leslie ldquoRegula-tion of signaling phosphoproteins by epidermal growth factorand Iressa (ZD1839) in human endometrial cancer cells thatmodel type I and II tumorsrdquoMolecular CancerTherapeutics vol4 no 12 pp 1891ndash1899 2005

[72] D Dai L Albitar T Nguyen L L Laidler M Singh and K KLeslie ldquoA therapeutic model for advanced endometrial cancersystematic progestin in combination with local adenoviral-mediated progesterone receptor expressionrdquo Molecular CancerTherapeutics vol 4 no 1 pp 169ndash175 2005


[73] G S Huang J Brouwer-Visser M J Ramirez et al ldquoInsulin-like growth factor 2 expression modulates taxol resistance andis a candidate biomarker for reduced disease-free survival inovarian cancerrdquo Clinical Cancer Research vol 16 no 11 pp2999ndash3010 2010

[74] G S Huang M J Gunter R C Arend et al ldquoCo-expression ofGPR30 and ER120573 and their association with disease progressionin uterine carcinosarcomardquo American Journal of Obstetrics andGynecology vol 203 no 3 pp 242e1ndash242e5 2010

[75] K K Leslie K W Thiel M J Goodheart K De Geest Y Jiaand S Yang ldquoEndometrial cancerrdquo Obstetrics and GynecologyClinics of North America vol 39 no 2 pp 255ndash268 2012

[76] S YangKWThiel andKK Leslie ldquoProgesterone the ultimateendometrial tumor suppressorrdquo Trends in Endocrinology andMetabolism vol 22 no 4 pp 145ndash152 2011

[77] C Kandoth N Schultz A D Cherniack et al ldquoIntegratedgenomic characterization of endometrial carcinomardquo Naturevol 497 no 7447 pp 67ndash73 2013


[79] P M Lenhart S Broselid C J Barrick L M Leeb-Lundbergand K M Caron ldquoG-protein-coupled receptor 30 interactswith receptor activity-modifying protein 3 and confers sex-dependent cardioprotectionrdquo Journal of Molecular Endocrinol-ogy vol 51 no 1 pp 191ndash202 2013

[80] F Gobeil Jr A Fortier T Zhu et al ldquoG-protein-coupledreceptors signalling at the cell nucleus an emerging paradigmrdquoCanadian Journal of Physiology and Pharmacology vol 84 no3-4 pp 287ndash297 2006

[81] M A Chotani and N A Flavahan ldquoIntracellular 1205722C-Ad-renoceptors storage depot stunted development or signalingdomainrdquo Biochimica et Biophysica Acta vol 1813 no 8 pp1495ndash1503 2011

[82] J Yu E Deliu X Q Zhang et al ldquoDifferential activationof cultured neonatal cardiomyocytes by plasmalemmal versusintracellular G protein-coupled receptor 55rdquo The Journal ofBiological Chemistry vol 288 no 31 pp 22481ndash22492 2013

[83] D Calebiro V O Nikolaev L Persani and M J Lohse ldquoSig-naling by internalized G-protein-coupled receptorsrdquo Trends inPharmacological Sciences vol 31 no 5 pp 221ndash228 2010

[84] G Sharma C Hu J L Brigman G Zhu H J Hathawayand E R Prossnitz ldquoGPER deficiency in male mice results ininsulin resistance dyslipidemia and a proinflammatory staterdquoEndocrinology 2013

[85] R L Brunsing K SOwens andE R Prossnitz ldquoTheGprotein-coupled estrogen receptor (GPER) agonist G-1 expands theregulatory T-cell population underTh17-polarizing conditionsrdquoJournal of Immunotherapy vol 36 no 3 pp 190ndash196 2013

[86] M R Meyer E R Prossnitz and M Barton ldquoThe G protein-coupled estrogen receptor GPERGPR30 as a regulator ofcardiovascular functionrdquo Vascular Pharmacology vol 55 no 1ndash3 pp 17ndash25 2011

[87] R Hammond D Nelson and R B Gibbs ldquoGPR30 co-localizeswith cholinergic neurons in the basal forebrain and enhancespotassium-stimulated acetylcholine release in the hippocam-pusrdquo Psychoneuroendocrinology vol 36 no 2 pp 182ndash192 2011

[88] D P Edwards ldquoRegulation of signal transduction pathways byestrogen and progesteronerdquo Annual Review of Physiology vol67 pp 335ndash376 2005

[89] E J Filardo and P Thomas ldquoMinireview G protein-coupledestrogen receptor-1 GPER-1 Its mechanism of action and rolein female reproductive cancer renal and vascular physiologyrdquoEndocrinology vol 153 no 7 pp 2953ndash2962 2012

[90] Q Wu K Chambliss M Umetani C Mineo and P W ShaulldquoNon-nuclear estrogen receptor signaling in the endotheliumrdquoThe Journal of Biological Chemistry vol 286 no 17 pp 14737ndash14743 2011

[91] E R Prossnitz and M Maggiolini ldquoMechanisms of estrogensignaling and gene expression via GPR30rdquo Molecular andCellular Endocrinology vol 308 no 1-2 pp 32ndash38 2009

[92] L Albitar G Pickett M Morgan J A Wilken N J Maihleand K K Leslie ldquoEGFR isoforms and gene regulation in humanendometrial cancer cellsrdquo Molecular Cancer vol 9 article no166 2010

[93] I Paterni S Bertini C Granchi M Macchia and F MinutololdquoEstrogen receptor ligands a patent review updaterdquo ExpertOpinion onTherapeutic Patents vol 23 no 10 pp 1247ndash1271

[94] F Minutolo M Macchia B S Katzenellenbogen and JA Katzenellenbogen ldquoEstrogen receptor 120573 ligands recentadvances and biomedical applicationsrdquo Medicinal ResearchReviews vol 31 no 3 pp 364ndash442 2011

[95] M RMeyer O Baretella E R Prossnitz andM Barton ldquoDila-tion of epicardial coronary arteries by the G protein-coupledestrogen receptor agonists G-1 and ICI 182780rdquo Pharmacologyvol 86 no 1 pp 58ndash64 2010

[96] I Mercier S Mader and A Calderone ldquoTamoxifen and ICI182780 negatively influenced cardiac cell growth via an estrogenreceptor-independent mechanismrdquo Cardiovascular Researchvol 59 no 4 pp 883ndash892 2003

[97] S L Palin P G McTernan L A Anderson D W Sturdee AH Barnett and S Kumar ldquo17120573-Estradiol and anti-estrogen ICIcompound 182780 regulate expression of lipoprotein lipase andhormone-sensitive lipase in isolated subcutaneous abdominaladipocytesrdquoMetabolism vol 52 no 4 pp 383ndash388 2003

[98] L Lam X Hu Z Aktary D W Andrews and M PasdarldquoTamoxifen and ICI 182780 increase Bcl-2 levels and inhibitgrowth of breast carcinoma cells by modulating PI3KAKTERK and IGF-1R pathways independent of ER120572rdquo Breast CancerResearch and Treatment vol 118 no 3 pp 605ndash621 2009

[99] M Singh G Setalo Jr X Guan D E Frail and C DToran-Allerand ldquoEstrogen-induced activation of the mitogen-activated protein kinase cascade in the cerebral cortex ofestrogen receptor-120572 knock-out micerdquo Journal of Neurosciencevol 20 no 5 pp 1694ndash1700 2000

[100] A H Kurt R N Tiftik I Un S Ulker and K Buyukafsar ldquoGprotein-coupled estrogen receptor1 (GPER1) maymediate Rho-kinase (ROCK-2) up-regulation in coronary endothelial cellsrdquoEndocrine Regulations vol 47 no 2 pp 75ndash84 2013

[101] P Diel ldquoTissue-specific estrogenic response and molecularmechanismsrdquo Toxicology Letters vol 127 no 1-3 pp 217ndash2242002

[102] V Craig Jordan J Lewis-Wambi H Kim et al ldquoExploiting theapoptotic actions of oestrogen to reverse antihormonal drugresistance in oestrogen receptor positive breast cancer patientsrdquoBreast vol 16 no 2 pp 105ndash113 2007

[103] G G J M Kuiper E Enmark M Pelto-Huikko S Nilssonand J-A Gustafsson ldquoCloning of a novel estrogen receptorexpressed in rat prostate and ovaryrdquo Proceedings of the NationalAcademy of Sciences of the United States of America vol 93 no12 pp 5925ndash5930 1996


[104] PHuang V Chandra and F Rastinejad ldquoStructural overview ofthe nuclear receptor superfamily insights into physiology andtherapeuticsrdquoAnnual Review of Physiology vol 72 pp 247ndash2722009

[105] V C Konigame E R Siu C Royer T F G Lucas C SPorto and F M F Abdalla ldquoEstrogen receptors mediate rapidactivation of phospholipase C pathway in the rat endometriumrdquoSteroids vol 76 no 14 pp 1582ndash1589 2011

[106] S Gingerich G L Kim J A Chalmers et al ldquoEstrogenreceptor alpha and G-protein coupled receptor 30 mediatethe neuroprotective effects of 17120573-estradiol in novel murinehippocampal cell modelsrdquo Neuroscience vol 170 no 1 pp 54ndash66 2010

[107] R Hammond R Mauk D Ninaci D Nelson and R B GibbsldquoChronic treatment with estrogen receptor agonists restoresacquisition of a spatial learning task in young ovariectomizedratsrdquo Hormones and Behavior vol 56 no 3 pp 309ndash314 2009

[108] S Liu G Kilic M S Meyers et al ldquoOestrogens improve humanpancreatic islet transplantation in a mouse model of insulindeficient diabetesrdquoDiabetologia vol 56 no 2 pp 370ndash381 2013

[109] S Patkar T D Farr E Cooper F J Dowell and H V OCarswell ldquoDifferential vasoactive effects of oestrogen oestrogenreceptor agonists and selective oestrogen receptor modulatorsin rat middle cerebral arteryrdquoNeuroscience Research vol 71 no1 pp 78ndash84 2011

[110] K M Sjoquist J Martyn R J Edmondson andM L Friedlan-der ldquoThe role of hormonal therapy in gynecological cancersmdashcurrent status and future directionsrdquo International Journal ofGynecological Cancer vol 21 no 7 pp 1328ndash1333 2011

[111] L M Berstein A E Tchernobrovkina V B Gamajunova etal ldquoTumor estrogen content and clinico-morphological andendocrine features of endometrial cancerrdquo Journal of CancerResearch and Clinical Oncology vol 129 no 4 pp 245ndash2492003

[112] C W Whitney V L Brunetto R J Zaino et al ldquoPhase II studyof medroxyprogesterone acetate plus tamoxifen in advancedendometrial carcinoma a Gynecologic Oncology Group studyrdquoGynecologic Oncology vol 92 no 1 pp 4ndash9 2004


and DNA mismatch repair genes Type 2 tumors are a het-erogeneous group of tumors including high-grade (undiffer-entiated) endometrioid carcinomas uterine papillary serouscarcinomas clear cell carcinomas and carcinosarcomas withdifferent mutational profiles Over 90 of uterine papillaryserous carcinomas are associated with p53 mutations 45ndash60 have Her-2neu mutations and PTEN mutations arerare [2 3] Carcinosarcomas which are characterized bymalignant epithelial andmesenchymal components are asso-ciated with many of the epidemiological risk factors linkedto endometrioid carcinomas including obesity and exposureto tamoxifen therapy suggesting that dysregulated estrogensignaling has a role in their pathogenesis and may representa therapeutic target Moreover recent mutational profilingstudies indicate that whereas some carcinosarcomas sharemutations with type 1 tumors (PTEN and ARID1A) othersshare mutations with uterine papillary serous carcinomas(notably p53 and PPP2R1A) [4]

The lack of estrogen receptor 120572 (ER120572) expression inmost type II tumors has led to the assumption that thesetumors must be ldquoestrogen-independentrdquo and that treatmentwith antiestrogens (selective estrogen receptor modulators(SERMs) such as tamoxifen and Raloxifene or pure antag-onistsselective estrogen receptor modulators (SERDs) suchas ICI182780fulvestrant) commonly used in breast can-cer treatment would be ineffectual a conclusion largelysubstantiated in a number of clinical trials [5 6] In factprolonged treatment of breast cancer with SERMs such astamoxifen leads to an increased incidence of endometrialcancer [7] particularly those of high-risk histologic types[8] resulting in significantly poorer overall survival [9] Theeffects of tamoxifen in the uterus have been ascribed toaltered expression of nuclear coregulatory proteins in theendometrium compared to the breast resulting in moderateagonist activity of tamoxifen in the uterus compared to itsantagonistic effects in the breast [10ndash13] However recentresults have suggested that a heretofore-underappreciatedestrogen receptor the G protein-coupled estrogen recep-tor (GPER formerly GPR30) may play an important rolein both the increased incidence of endometrial cancer inwomen treated with tamoxifen [14] as well as representingan alternate mechanism through which endometrial cancersparticularly type II tumors can maintain responsiveness toestrogen [15]

GPER is a member of the 7-transmembrane spanningG protein-coupled receptor (GPCR) superfamily struc-turally unrelated to the nuclear receptor family membersER120572 and ER120573 [16 17] Activation of GPER by estrogenhas been demonstrated in many cancer cell lines [18 19]including endometrial cancer cells [15 20ndash27] GPER isalso activated by antiestrogens including tamoxifen (ie 4-hydroxytamoxifen) [28] and ICI182780 (fulvestrant) [29]leading to the suggestion that GPER plays a role in hormone-resistance in breast cancer [30 31] as well as in theincreased incidence of endometrial cancer in women takingtamoxifen for breast cancer [14 32] Furthermore GPER(over)expression has been associated with many cancersand in particular poor prognosis in a number of cancersincluding breast [33] ovarian [34] lung [35] pancreatic [36]

and endometrial [37] although observations to the contraryhave also been reported [38 39]

Because of the lack of specificity of estrogen and anti-estrogens for the three known estrogen receptors (ER120572 ER120573andGPER)we have developed both aGPER-selective agonist(G-1 [40]) and antagonists (G15 [41] and G36 [42]) thatdisplay virtually no activity towards the classical estrogenreceptors In the current study we examine the expressionand function of GPER in the ER120572minusER120573minus endometrial cancercell line Hec50 [43 44] which is representative of type IIendometrial cancers We demonstrate that in Hec50 cellsGPER is localized predominantly in intracellular membranesand mediates PI3K and ERK activation in response to bothestrogen and G-1 as well as the antiestrogens tamoxifenRaloxifene and ICI182780 and the ldquoER120572-selectiverdquo agonistpropylpyrazole triol (PPT) We also demonstrate that Hec50cells and primary patient endometrial adenocarcinomasmaintain expression of GPER when grown as xenografttumors Finally we demonstrate both estrogen and G-1stimulate Hec50 xenograft tumor growth in vivo and that theGPER antagonist G36 greatly reduces growth of estrogen-stimulated Hec50 tumors Overall these results suggest thatGPER may play a critical role in endometrial carcinogenesisproviding a novel target for prognosis and treatment

2 Materials and Methods

Reagents 17120573-estradiol 17120572-estradiol 4-hydroxytamoxifenRaloxifene genistein LY294002 bovine serum albumin(BSA) normal goat serum insulin transferrin hydrocor-tisone fetuin pancreatin and trypsin were from Sigma(St Louis MO USA) AG1478 and GM6001 were fromCalbiochem (BillericaMA USA) DPN PPT and ICI182780were from Tocris Chemicals (Ellisville MO USA) G-1G15 and G36 were synthesized as previously described [40ndash42] Goat anti-rabbit Alexa-488 goat anti-rabbit Alexa-568and donkey anti-mouse Alexa-568-conjugated secondaryantibodieswere from Invitrogen (Carlsbad CAUSA) Rabbitanti-GPER C-terminal antiserum (cross-reactive to bothhuman and murine sequences) was produced as previouslydescribed and used at a dilution of 1 10000 [28] Rab-bit anti-GPER human N-terminal antiserum was producedagainst the peptide sequence MDVTSQARGVGLEMYPG-TAQPAAC (with an added carboxy-terminal cysteine forconjugation to KLH) by New England Peptide Inc (GardnerMA USA) and used at a dilution of 1 5000 Polyclonalantibodies against total ERK and pERK were from Cell Sig-naling (Danvers MA USA) monoclonal anti-actin and anti-120573-catenin antibodies were from Millipore (Burlington MAUSA) Goat anti-rabbit HRP and donkey anti-mouse HRPwere from GE-Amersham (Piscataway NJ USA) DulbeccorsquosMEM RPMI 1640 and phenol red-free DMEMF12 mediapenicillinstreptomycinglutamine and fetal bovine serumwere obtained from Fisher (Pittsburgh PA USA)

Cell Culture andTransfection Human endometrial carcinomaIshikawa H cells and Hec50 (specifically Hec50co [44]) cells(kindly provided by K K Leslie) were cultured in DMEMmedium with FBS (10) and 100UmL penicillin 100 120583gmL





antibiotics









Hec50

H cell


(a)

GPER

Actin

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(b)

lowast

Relat

ive e

xpre

ssio

n (

)

0

20

40

60

80

100

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(c)








3 Results




N-Term

C-Term

GPER 120573-catenin









17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)








nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References


























































































































antibiotics









Hec50

H cell


(a)

GPER

Actin

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(b)

lowast

Relat

ive e

xpre

ssio

n (

)

0

20

40

60

80

100

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(c)








3 Results




N-Term

C-Term

GPER 120573-catenin









17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)








nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References























































































































Hec50

H cell


(a)

GPER

Actin

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(b)

lowast

Relat

ive e

xpre

ssio

n (

)

0

20

40

60

80

100

H cell Hec50

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

Unt

rans

fect

ed

Con

trol s

iRN

A

GPE

R siR

NA

(c)








3 Results




N-Term

C-Term

GPER 120573-catenin









17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)








nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References
























































































































N-Term

C-Term

GPER 120573-catenin









17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)








nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References
























































































































17120573E2 + LY 17120573E2 + AG 17120573E2 + GM

(a)

10nMG-1 G15

100nM G15 100nM G15100nM+ 10nM 17120573E2 +10nM G-1

G36100nM G36 100nM G36100nM

+ 10nM 17120573E2 +10nM G-1

G-1 + LY G-1 + AG G-1 + GM

(b)








nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References























































































































nstim

10nM

E2

10nM

G-1


(a)


GPE

R

(b)











(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References

























































































































(a)

pERK


(b)

lowast


lowast

lowast

lowast


lowastlowast

170

130

100

75

50

25

0

pERK

()

DM

SO

Estro

gen

G-1

4O

H-ta

mox

ifen

Ralo

xife

ne

Gen

istei

n

PPT

DPN EG

F

Stimulus

ICI182

780

+1120583M G15

(c)





(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References























































































































(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)





800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References























































































































800

700

600

500

400

300

200

100

0

Tum

or v

olum

e (m

m3)

0 3 6 9 12 15 18 21 24

E2G-1

E2 + G36

Sham

lowast

lowast

Time (days)

(a)

E2 G-1 E2 + G36Sham

lowast

lowast

Tum

or m

ass (

g)

06

05

04

03

02

01

00

(b)






4 Discussion



















5 Conclusions




Acknowledgments



References







































































































































5 Conclusions




Acknowledgments



References






















































































































g protein-coupled estrogen receptor-selective ligands modulate

Documents